539708 PSPXXX10.1177/0146167214539708Personality and Social Psychology BulletinKteily et al research-article2014 Article “Not One of Us”: Predictors and Consequences of Denying Ingroup Characteristics to Ambiguous Targets Personality and Social Psychology Bulletin 2014, Vol 40(10) 1231–1247 © 2014 by the Society for Personality and Social Psychology, Inc Reprints and permissions: sagepub.com/journalsPermissions.nav DOI: 10.1177/0146167214539708 pspb.sagepub.com Nour Kteily1, Sarah Cotterill2, Jim Sidanius2, Jennifer Sheehy-Skeffington2, and Robin Bergh3 Abstract We investigated individual difference predictors of ascribing ingroup characteristics to negative and positive ambiguous targets Studies and investigated events involving negative targets whose status as racial (Tsarnaev brothers) or national (Woolwich attackers) ingroup members remained ambiguous Immediately following the attacks, we presented White Americans and British individuals with the suspects’ images Those higher in social dominance orientation (SDO) and right-wing authoritarianism (RWA)—concerned with enforcing status boundaries and adherence to ingroup norms, respectively—perceived these low status and low conformity suspects as looking less White and less British, thus denying them ingroup characteristics Perceiving suspects in more exclusionary terms increased support for treating them harshly, and for militaristic counter-terrorism policies prioritizing ingroup safety over outgroup harm Studies and experimentally manipulated a racially ambiguous target’s status and conformity Results suggested that target status and conformity critically influence SDO’s (status) and RWA’s (conformity) effects on inclusionary versus exclusionary perceptions Keywords group categorization, ingroup overexclusion, hypodescent, social dominance orientation, right-wing authoritarianism, terrorism Received October 10, 2013; revision accepted May 21, 2014 On the 15th of April 2013, two bombs exploded near the finish line of the Boston Marathon, killing people and injuring 280 others Surveillance footage quickly determined two primary suspects, captured in grainy photos wearing backpacks near the incident Speculation immediately abounded about their background and potential motivations, fueled by the ambiguity of their skin color and the difficulty in determining whether the incident was carried out by foreign or American actors Shortly thereafter, the FBI released the names of Tamerlan and Dzhokhar Tsarnaev Unusually, although we quickly learned more about their life histories, many of the outstanding questions about the motivation behind the attacks, and the racial and ethnic status of the perpetrators, remained unanswered (Walsh, 2013) A complex picture emerged: Although their names clearly sounded foreign to most Americans, we learned that they had been living in the United States for many years Moreover, in spite of the fact that they did not quite fit the physical profile and background of the average White person familiar to Americans, their ethnic background was from the North Caucasus, the very region that gave name to the term “Caucasian.” In light of their ambiguous background, the question of whether or not the bombers were White became highly salient and hotly debated by social commentators and media members For example, David Sirota wrote in Salon the day after the attacks (prior to their identities being known) about his hope that the terrorists were White and homegrown He argued that if the bombers turned out to be White, they would be perceived as an aberration or an anomaly and would not provoke aggressive responses toward other groups by the U.S government (Sirota, 2013) This prompted significant backlash among individuals bothered by the assertion that a White individual could be capable of these attacks (Walsh, 2013) Nevertheless, even after the images of the brothers were released, their racial identity was the subject of much debate, as evidenced by headlines such as “Are the Tsarnaev Brothers White?” (Walsh, 2013), “Are the Kellogg School of Management, USA Harvard University, Cambridge, MA, USA Uppsala University, Sweden Corresponding Author: Nour Kteily, Kellogg School of Management, USA Email: n-kteily@kellogg.northwestern.edu Downloaded from psp.sagepub.com at Harvard Libraries on February 18, 2015 1232 Personality and Social Psychology Bulletin 40(10) Tsarnaevs White?” (Beinart, 2013), and “The Unbearable Whiteness of the Brothers Tsarnaev” (First Post, 2013) In sum, the potential Whiteness of the Tsarnaev brothers became a highly salient dimension in the aftermath of the Marathon attacks and one presumably highly relevant to White Americans.1 Shortly after the attack in Boston, another incident occurred in Woolwich, a working-class neighborhood in Southeast London, the United Kingdom In this case, the suspects—Michael Adebolajo and Michael Adebowale— were racially unambiguous Nevertheless, in the immediate aftermath of their suspected lethal attack on a British soldier, their national status (as British citizens or as foreigners) remained unclear Thus, although there were important differences in the nature and context of the terrorist attacks, an important and unusual similarity between these two events was the fact that the status of the perpetrators as members of individuals’ own group or members of an external group was ambiguous These two incidents provided naturalistic settings in which to investigate an important question: namely, the variables predicting the perception of ambiguous targets in ingroup versus outgroup terms Indeed, the issue of imbuing targets with ingroup characteristics is one with important consequences, both for the individuals being characterized and for the groups involved Research in intergroup relations has long shown that categorization into groups results in a preference for one’s own group over the outgroup (Allport, 1954; Gramzow & Gaertner, 2005; Sherif, 1967) This bias manifests itself in terms of increased positive regard and favoritism toward the ingroup relative to the outgroup (Mullen, Brown, & Smith, 1992; Tajfel & Turner, 1986), increased empathy and prosocial behavior toward members of the ingroup (Hornstein, 1976; Piliavin, Dovidio, Gaertner, & Clark, 1981), construal of other ingroup members as closer to the self (Turner, Hogg, Oakes, Reicher, & Wetherell, 1987), and increased levels of trust and cooperation (Miller, Maner, & Becker, 2010) In sum, group members’ orientation toward other individuals is importantly influenced by whether they perceive these individuals in ingroup versus outgroup terms The determination of who belongs to the ingroup also matters for the group itself Individuals are motivated to hold their group in positive regard (Tajfel & Turner, 1986) and are concerned with its level of overall functioning, as well as its status and standing in society (Stelzl, Janes, & Seligman, 2007) These concerns should influence the ascription of ingroup membership to others First, we would expect individuals to be sensitive to how those they admit to the ingroup influence its operation Thus, they should attempt to ensure that those ascribed ingroup status will conform to the norms and practices of the group and support its smooth functioning and cohesiveness (Feldman, 1984) Second, we expect that group members will employ strategies to enhance and protect its image: One such strategy is to ascribe group membership to valued individuals and deny it to undesirable others (Castano, Yzerbyt, Bourguignon, & Seron, 2002; Leyens & Yzerbyt, 1992) Examples of this include the conferral of honorary degrees to venerated individuals by universities, and companies dropping celebrity endorsees who engage in egregious, socially undesirable behavior from their “corporate family.” Consistent with the notion that group members are concerned with the consequences for their group of granting membership to others, Castano and colleagues (2002) argued that individuals take care when making ingroup categorizations Such caution helps avoid the “contamination” of the ingroup and any damage to its functioning or status that might ensue (see also Ho, Sidanius, Cuddy, & Banaji, 2013; Leyens & Yzerbyt, 1992; Stelzl et al., 2007) At the same time, it may pay not only to exclude negative ambiguous targets that may contaminate the group but also to include positive ambiguous targets and thus to “bask in their reflected glory.” Indeed, consistent with both exclusion of negative and inclusion of positive ambiguous targets, Stelzl et al (2007) found that Canadians were more likely to see Ben Johnson (a Jamaican-born Canadian sprinter) as Canadian after he won the gold medal at the 1988 Olympics, but as Jamaican after he was subsequently disqualified for steroid use Individual Differences in the Conferral of Ingroup Characteristics Although our reasoning suggests that all group members should show some concern with determining who does and who does not belong to their group, there is nevertheless theoretical reason to expect individual differences in how discriminating individuals are in ascribing others with ingroup characteristics, a question that has received scant empirical investigation (but see Blascovich, Wyer, Swart, & Kibler, 1997; Ho et al., 2013; Krosch, Berntsen, Amodio, Jost, & Van Bavel, 2013) In this work, we considered the role of two individual difference variables—social dominance orientation (SDO; Pratto, Sidanius, Stallworth, & Malle, 1994) and right-wing authoritarianism (RWA; Altemeyer, 1981)—that both relate to a wide variety of socio-political attitudes and behaviors (e.g., racism, sexism, support for war, support for the death penalty; Altemeyer, 1981; Kteily, Ho, & Sidanius, 2012) Although these variables predict prejudice toward similar groups in practice, they so independently of one another and for unique reasons (Duckitt, 2001) Whereas being high in RWA reflects a concern with traditionalism, submission to established authorities, and aggression toward those who violate the social norms of the ingroup, individuals high in SDO favor the maintenance of anti-egalitarian, hierarchical relationships between social groups, and the domination of “inferior” groups by “superior” ones As such, when it comes to perceiving ambiguous Downloaded from psp.sagepub.com at Harvard Libraries on February 18, 2015 1233 Kteily et al targets in inclusionary versus exclusionary terms, it is important to consider the role of both ideological orientations rather than either one alone Individuals high in RWA should be especially sensitive to the conformity of ambiguous targets in their ascription of ingroup characteristics As such, we expected individuals high in RWA to see nonconformist individuals as potential threats to group cohesion and thus to be more willing to perceive them in exclusionary terms (Duckitt, 2001; Thomsen, Green, & Sidanius, 2008) However, if our expectation about the role for ambiguous target conformity is correct, such exclusionary perceptions may not be extended by high RWA individuals to more conformist ambiguous targets Unlike RWA, individuals high in SDO should be primarily concerned with the status of ambiguous targets, given high SDO individuals’ concerns about maintaining and reifying group status boundaries (Ho et al., 2013) Thus, individuals high in SDO should tend to exclude low status ambiguous targets who they might perceive as threats to the status boundary between their group and inferior groups In contrast, including higher status ambiguous targets should not blur (and may in fact sharpen) group status boundaries, and thus, we would not expect high SDO individuals to see such targets in exclusionary terms Whereas a few studies (Ho et al., 2013; Krosch et al., 2013) have investigated a role for SDO in the realm of hypodescent research (the tendency to categorize half-racials, a specific case of ambiguous target, as members of the low status group), no studies have investigated the relationship between RWA and the ascription of ingroup characteristics to ambiguous targets Finally, no studies have looked at individual differences in the ascription of ingroup characteristics to positive ambiguous targets The Present Work We tested our ideas across four studies In Studies and 2, we assessed the role of SDO and RWA in the ascription of ingroup characteristics to low status and low conformity (i.e., negative) ambiguous suspects in the immediate aftermath of terrorist attacks By collecting data within days of the attacks, we were able to examine our hypotheses in the context of highly salient real-world incidents By assessing two different contexts, and two different bases of ingroup membership, we were also able to investigate the generalizability of our hypotheses Thus, we examined the role of SDO and RWA in influencing perceptions of the Whiteness of the racially ambiguous Tsarnaev brothers (Study 1), as well as investigating perceptions of the Britishness of Michael Adebolajo (Study 2), one of the suspected Woolwich attackers These targets’ actions were perceived, in no uncertain terms, to be extremely low in status and conformity Because individuals high in RWA tend to reject nonconformist behavior that challenges established authorities and threatens group cohesion, we hypothesized that White individuals high on RWA would downplay the “Whiteness” of the Tsarnaev brothers, and British individuals high on RWA would similarly downplay the “Britishness” of the Woolwich suspects Because individuals high in SDO seek to avoid blurring group status boundaries, we also expected that high SDO would be associated with exclusionary perceptions of the Tsarnaev brothers and the Woolwich suspects Distancing them from the ingroup in this way allows individuals high in SDO to avoid having the status of their group “contaminated” by an association with the extremely low status targets Because RWA and SDO have been frequently observed to operate in parallel, we expected each construct to contribute uniquely to exclusionary perceptions targeted at the negative ambiguous targets In Studies and 4, we sought to extend our examination to more positive ambiguous targets and in so doing, assess our claims about the reasons for SDO’s and RWA’s effects Specifically, we experimentally manipulated the status (Study 3) and conformity (Study 4) of a racially ambiguous target to be low or high Consistent with our expectation that individuals high in SDO would be influenced by target status, we expected ambiguous targets’ status to moderate the effect of high SDO (but not high RWA) on individuals’ ascription of ingroup characteristics Consistent with our expectation that individuals high in RWA would be influenced by target conformity, we expected high RWA (but not high SDO) individuals’ ascription of ingroup characteristics to be moderated by information about how conformist versus nonconformist the ambiguous target was Our central theoretical interest in this work was determining the predictors of perceiving ambiguous target in ingroup terms Nevertheless, we also expected that ingroup characteristics—once ascribed or denied—would have important consequences We assessed these ideas in our first two studies We expected the perception of the Tsarnaev brothers and the Woolwich attackers in ingroup versus outgroup terms to matter Previous work has described the benefits ingroup members receive from their membership within the group (e.g., trust: Brewer, 2008; empathy: Piliavin et al., 1981; altruism: Stürmer, Snyder, Kropp, & Siem, 2006) However, being a member of an outgroup subjects one to the potential application of any of a number of processes—such as stereotyping, prejudice, and dehumanization—that can be used to justify aggressive attitudes and behavior, much more difficult to justify toward members of the ingroup Thus, we hypothesized that the perception of the Tsarnaev brothers and Woolwich attackers as outgroup (as opposed to ingroup) members would be associated with harsher judgments about the punishments they deserve.2 Along these same lines, characterizing negative ambiguous targets who have harmed the group as outgroup members might contribute to the justification of aggressive policies toward outsiders more generally Perceiving the ingroup to have been targeted from the outside may increase group members’ support for policies and institutions designed to Downloaded from psp.sagepub.com at Harvard Libraries on February 18, 2015 1234 Personality and Social Psychology Bulletin 40(10) protect the ingroup, at the expense of outsiders, shifting group members’ moral calculus to further prioritize ingroup over outgroup outcomes (Stürmer, Snyder, Kropp, Siem, & Kiel, 2006; see also Lickel, Miller, Stenstrom, Denson, & Schmader, 2006) As such, we hypothesized that perceptions of the Tsarnaev brothers as less White and the Woolwich attackers as less British would relate to support for aggressive counter-terrorism measures: measures with the stated aim of protecting the ingroup, but that nevertheless had the potential for grave consequences and negative outcomes for those not belonging to the group (see also Asbrock & Fritsche, 2013) Study In the first study testing these hypotheses, we conducted a two-wave survey with White American participants in the direct aftermath of the Boston Marathon attacks Two days following the attack, we assessed, among other things, the individual difference variables of interest and demographics We then followed up with participants approximately 10 days later In this second wave, we measured perceptions of the Whiteness of the Tsarnaev brothers, a construct that we expected would be salient to White participants given the heated debates about the brothers’ racial background Our assessment of Whiteness perceptions was based on a purely perceptual measure, with participants rating how White the Tsarnaev brothers looked in the photos released by the FBI At the second wave, we also assessed their support for punishment of the Tsarnaev brothers and for militaristic counterterrorism policies Method Participants Five hundred seventy-four participants completed Wave of the study (52.2% female; M age = 32.45 years, SD = 10.84 years) Of these participants, we selected only White Americans (n = 426 participants) Thirty participants were excluded because they reported a score less than on a to scale asking participants to indicate how seriously they took the study at Wave Thus, our final Wave sample was 396 White American participants (51.8% female, M age = 33.53, SD = 11.24) Of these participants, 261 (65.9%) also completed Wave of the survey Ten further participants were excluded from Wave analyses on the basis of the same seriousness check at Wave Thus, the final Wave sample was 251 White American participants (54.0% female; M age = 34.11, SD = 11.56).3,4 Data were collected using Amazon’s Mechanical Turk platform, as part of a broader data collection effort on attitudes toward the Boston Marathon attacks The first wave was administered at 5:00 p.m on April 18, 2013 At this time, pictures of the suspects had just been released by law enforcement officials, but their names, identities, and backgrounds were still unknown (as remained the case until completion of Wave data collection) Data collection was completed within a few hours, ensuring that all participants had essentially the same amount of information about the events and the suspects The second wave of the study was launched on April 26, 2013, after the first suspect had been killed, and the second arrested Data collection for this wave was terminated on May 1, 2013 Measures Wave variables Demographics We assessed participants’ age and gender (1 = male; = female) We assessed social class by asking them to answer the following question: “How would you describe your family’s social class position?” (1 = poor; = working class; = middle class; = upper-middle class; = upper class) We assessed their level of education by asking them to indicate the highest level of education they had completed (1 = no formal education; = elementary school; = some high school; = completed high school; = some college; = BA/BS degree; = some graduate/professional school; = hold graduate/professional degree) SDO was measured using randomly selected, counterbalanced, items from the 16-item SDO-6 scale (Pratto et al., 1994) Sample items include the following: “It’s OK if some groups have more of a chance in life than others,” and “No one group should dominate in society” (reverse-coded) RWA was assessed with 12 items taken from Altemeyer’s (1981) scale Sample items include the following: “People should pay less attention to the Bible and other old traditional forms of religious guidance and instead develop their own personal standards of what is moral and immoral” (reverse-coded) and “Young people sometimes get rebellious ideas, but as they grow up they ought to get over them and settle down.” Experimental condition At the end of Wave 1, participants were asked to read a text (see the online appendix) that served as an experimental manipulation for purposes unrelated to the current study In one condition, participants read a text arguing that Americans should consider the effects of America’s own policies on the world rather than only considering the costs of terrorism to Americans In a second condition, participants read a text arguing that America was facing increasing threats to its security In a third (control) condition, participants read no text We included experimental condition as a control variable in all analyses.5 Wave variables Whiteness perceptions Perceptions of the Whiteness of the Tsarnaev brothers was measured at Wave and assessed using four items Participants were shown the two sets of pictures of the Tsarnaev brothers released by law enforcement officials For each set of pictures, participants read, Downloaded from psp.sagepub.com at Harvard Libraries on February 18, 2015 1235 Kteily et al Above is a photograph released by the FBI on Thursday, April 18th of the lead suspects in the Boston Marathon bombing investigation On the left is Dzhokhar Tsarnaev and on the right is Tamerlan Tsarnaev How White you think the suspects look? Use the slider to indicate where you think each of the suspects falls on a continuum from Non-White to White For each of the brothers, participants indicated their Whiteness perceptions using a 100-point slider scale, where “0” indicated “Non-White” and 100 indicated “White.” Thus, participants completed four slider scale items in total (once for each brother in each picture) We averaged these items to create our index of Whiteness perceptions Aggressive responses to the marathon attacks Hrsh treatment Like Whiteness perceptions, this construct was also assessed at Wave First, we assessed participants’ responses to six items: “The perpetrator of the Boston Marathon attacks deserves to die as painful a death as possible,” “I hope the perpetrator of the Boston Marathon attacks rots in hell,” “The perpetrator of the Boston Marathon attacks is entitled to the best legal counsel available” (reverse-coded), “We shouldn’t rush to judgment in bringing the perpetrator of the Boston Marathon attacks to justice” (reverse-coded), “It is OK for Tsarnaev not to have been read his Miranda rights before interrogation,” and “It is appropriate to charge Tsarnaev with the use of a weapon of mass destruction.” Participants indicated their responses to each item using a 7-point scale, where indicated strongly disagree and indicated strongly agree Second, participants were asked the following question: If found guilty of planning and executing the Boston marathon attack, I would recommend that Tsarnaev be sentenced with: = Maximum of 20 years in prison with the possibility of parole; = 20-40 years in prison with the possibility of parole; = Life in prison with the possibility of parole; = Life in prison without the possibility of parole; =The death penalty Because the final item was on a different scale from the remaining items, we standardized all scores to compute the composite index of harshness Militaristic counter-terrorism This construct was measured at Wave by asking participants to rate their agreement with each of the following 11 items: “To put an end to terrorist acts, I think it is OK to use enhanced interrogation techniques”; “To put an end to terrorist acts, I think it is OK to use torture”; “To put an end to terrorist acts, I think it is OK to use waterboarding”; “To put an end to terrorist acts, I think it is OK to target civilians and combatants alike in foreign terrorist strongholds”; “To put an end to terrorist acts, I think it is OK to bomb an entire country if it is known to harbor anti-American terrorists”; “To put and end to terrorist acts, I think it is OK to target Muslims with extra profiling and surveillance”; “I support the war in Afghanistan”; “I support continued military efforts abroad to root out potential terrorists”; “We should spend more time on diplomatic efforts as opposed to engaging in military activity abroad” (reversecoded); “We shouldn’t be afraid to hunt down anyone who threatens our country anywhere”; and “We should strike back with brutal force against anyone who seeks to intimidate us.” Participants responded to each item using a 7-point Likert-type scale, where indicated strongly disagree and indicated strongly agree Results and Discussion Descriptive statistics We report descriptive statistics, scale reliabilities, and variable intercorrelations in Table We were particularly interested in participants’ perceptions of the Whiteness of the Tsarnaev brothers Consistent with our notion that there was some ambiguity surrounding their racial group membership, the average Whiteness rating in our sample was only slightly above the midpoint, with substantial individual variability In this work, we were primarily interested in assessing the predictors of the ascription of ingroup characteristics Specifically, we predicted that individuals higher in RWA and individuals higher in SDO would be less likely to ascribe ingroup characteristics to the racially ambiguous Tsarnaev brothers As such, we included each of these variables as a predictor of our index of Whiteness perceptions in a simultaneous regression, controlling for demographic variables (age, gender, class, and education) and experimental condition As expected, we found that each of RWA (b = −4.33, β = −.19, p = 01, 95% confidence interval [CI] = [−.05, −.33]) and SDO (b = −6.71, β = −.30, p < 001, 95% CI = [−.16, −.43]) was uniquely associated with participants reporting that the Tsarnaev brothers looked less White in the photographs they rated We also observed that age was associated with increased Whiteness perceptions (b = 0.28, β = 12, p = 04, 95% CI = [.004, 24]), though it was uncorrelated with Whiteness in zero-order terms A secondary question concerned how perceptions of the Whiteness of the Tsarnaev brothers would influence the harshness of attitudes toward them and might influence the support for aggressive counter-terrorism policies that prioritized the safety of the ingroup at the potential expense of outsiders We were interested in whether Whiteness perceptions would predict these outcomes uniquely over and above the other variables we were investigating As such, we first included each of RWA, SDO, the demographic controls, and experimental condition in the first step of a hierarchical regression, and then added Whiteness perceptions at the second step As can be seen in Table 2, RWA and SDO emerged as significant predictors of greater harshness at Step Nevertheless, adding Whiteness perceptions at the second step significantly increased the predicted proportion of vari2 ance in harshness, Rchange = 02, F(1, 237) = 8.12, p = 005 Downloaded from psp.sagepub.com at Harvard Libraries on February 18, 2015 1236 Personality and Social Psychology Bulletin 40(10) Table Descriptive Statistics and Intercorrelations of Variables Assessed in the Aftermath of the Boston Marathon Attacks Variable SDO RWA Whiteness Harsh treatment Militaristic counter-terrorism Age Gender Class Education M SD α Number of items — 46*** −.37*** 42*** 56*** −.04 −.10 18** −.05 2.43 1.19 90 — −.33*** 60*** 65** 13* 21** 04 05 3.61 1.17 88 12 — −.37*** −.37*** 09 −.08 -.00 −.05 63.57 26.77 93 — 72*** 07 13* 07 02 -.01 0.70 82 — 10 −.01 15* 01 3.25 1.42 92 11 — 12* −.05 07 33.91 11.50 — — −.02 06 — — — — 03 2.66 0.75 — — 4.24 1.80 — Note SDO = social dominance orientation; RWA = right-wing authoritarianism *p < 05 **p < 01 ***p < 001 Table Hierarchical Regression Predicting White American Support for Harsh Treatment of the Tsarnaev Brothers in the Aftermath of the Boston Marathon Attacks β 95% CI β p Step (R = 39) SDO RWA Age Gender Education Class Condition Step (R2 = 41) SDO RWA Age Gender Education Class Condition Whiteness Table Hierarchical Regression Predicting White American Support for Militaristic Counter-Terrorism Policies in the Aftermath of the Boston Marathon Attacks 95% CI p 19 51 01 04 −.01 02 −.02 [.07, 30] [.39, 63] [−.09, 11] [−.07, 14] [−.11, 10] [−.08, 12] [−.12, 08] 002